Spinal mu, delta and alpha2, muscarinic, GABA-B and adenosine Al agonists produce a powerful analgesia in animals and, where examined, in humans. Despite their distinct receptors, they have several common properties. I. Except for A1/GABA-B sites, all exert a clear presynaptic effect inhibiting release of small afferent transmitters; and, hyperpolarize cells. 2. Continued spinal delivery of the agent (toleragen) results in reduction in effect otherwise produced by a given dose and a right shift in the "probe" or test drug response curve. 3. Injection of the respective antagonists evokes a withdrawal state characterized by a spinal hyperreflexia, hyperalgesia and allodynia. Co-delivery of NMDA antagonists can minimize the reduction in the effect produced by morphine and attenuate naloxone-induced withdrawal. In recent work, we have shown a similar interaction of spinal NMDA antagonists with chronic opiates resulting in a sustained opiate action, prevention of the right shift in spinal probe dose response curve and suppression of withdrawal. These observations jointly raise specific hypotheses. l) If the NMDA/non-NMDA or metabotrophic receptor are involved, then a variety of the respective antagonists should be effective, with their relative potency being related to their affinity for the binding site, whether it be at the receptor, the channel, or NMDA-glycine site. 2) If [Ca]- plays a role, non-NMDA/metabotrophic site antagonists may reduce tolerance evolution. Agents which reduce voltage-sensitive Ca flux (as do N, P and L channel antagonists) would be anticipated to have similar actions. 3) Spinal NMDA receptor activation yields an increase in spinal NO release and systemic NOS inhibitors have been shown to diminish systemic tolerance, then spinal delivery of NOS synthase inhibitors should yield similar effects. Similarly, the potential role of phosphorylating enzymes activated by increased Ca would be revealed by protein kinase C inhibitors. 4) Spinal receptor systems outlined above do not show a significant heterologous cross-tolerance, but several exert a similar spinal action (presynaptic inhibition of C fiber release and hyperpolarization of dorsal horn neurons by a Gi/o protein-mediated increase in K conductance). We hypothesize that NMDA antagonists will prevent evolution of the tolerant state and the withdrawal associated with chronic delivery of these agents. 5) Finally, it has been argued that chronic pain attenuates tolerance onset. This speculation is contrary to the fact that pain states reflect enhanced glutamate receptor activity. Spinal tolerance will be studied in rats prepared with nerve injury induced allodynia or hyperalgesia, or an inflammatory pain state or a hyperpathic state induced by chronic spinal NMDA infusion. To address these issues, we will employ the well defined rat and dog chronic intrathecal infusion model wherein the toleragen alone or together with adjuvant agents will be infused and the effect upon a thermal escape latency assessed. Aside from the clear relevance to mechanisms, these results have a correspondingly important clinical significance in the management of chronic pain states.